The definition of the immunity level of an electronic device is based on the realization of EMC immunity test. A specific power over a wide range of radio frequencies (RF) is applied to the device, verifying whether the device continues to function properly. Depending on the range of RF, the type of testing can be classified as conducted (150 kHz up to 80 MHz) where the power is conducted into equipment under test (EUT) via cables. It can also be classified as radiated (80 MHz up to 6 GHz) where the power is radiated over-the-air through an antenna. In this regard, depending on frequency, distance, and specified antenna, an output power of hundreds or thousands of watts is required to generate field strength of a given intensity at a given distance from the antenna. But, most of the signal generators are not capable of producing such an amount of power. Therefore, an EMC amplifier must be considered to perform the testing.
An amplifier is an active device designed to increase the amplitude of the input signal, producing an output signal with the same shape and frequency characteristics as the input signal. They are important equipment of the signal chain of the EMC testing, specially for radiated immunity testing, that often requires high electric field strengths ranging from 3 to 200 V/m. The field strength is defined as a function of several variables plus the power needed to create the expected field strength, which is frequency-dependent. In this sense, the characteristics of the amplifier become so important, as it defines how much power can be delivered to generate the E-field over the frequency range of interest. The amplifier must be able to generate sufficient output power over the range of frequency during the EMC testing, but for broadband amplifiers, the output power shows variation over its operating frequency range. In regard to its operational characteristics, amplifiers operate on its linear region when there is a fixed increase of the output signal amplitude level in respect to the input signal (measured in dB). Although, when linear relationship between input and output power is no longer available, in other words, increasing the delivered power no more increases the amplitude of the output signal, it is said that the amplifier reached its compression point. The 1-dB compression point is the metric to define the linearity of a given amplifier. It is a measure of the amplifier’s ability to handle input signal power before the output signal starts to compress. At this point, the output power of the amplifier is 1 dB below its input power. Another metric used to gauge the linearity of an amplifier is the third-order intercept point (IP3) that represents the theoretical output power level, where the power of third-order intermodulation products (IM3) are equal to the fundamental output signals. Amplifiers operating close to its compression can generate harmonics and intermodulation products (IM) at frequencies other than the fundamental frequency it should operate. These undesired products can increase the power level of the output signal, but in frequencies other than the desired frequency, which is not desirable during the EMC immunity test.
However, what can happen if the amplifier behaves non-linearly, and possibly generates frequencies other than the frequency of interest? How to make sure that the EUT is exclusively reacting to the energy at the intended frequency? The existence of harmonics and intermodulation products can compromise the identification of the specific frequency components that cause malfunctions in the EUT [2] [3] [4]. The harmonics power is low compared to the power of the fundamental frequency, but, for instance, it can yield a higher field strength due to the frequency response of the antenna. Therefore, it is essential to avoid harmonics during the EMC testing. In the case of IM, that are generated when using multi-tone testing signals, these components are found not only at harmonic frequencies, but also at the sum and difference frequencies of the original frequencies and at multiples of those frequencies. EMC test standards define harmonics levels that can be accepted during testing. And, particularly, it guides the prevention of intermodulation products by using power back-off to keep the amplifier as linear as possible, which means to let the amplifier operate within its linear region. The other way is to use an amplifier with a 1 dB compression point above the maximum output power level. However, an amplifier that operates exclusively within the linear region is not highly efficient, and achieving high efficiency involves operating the PA in a nonlinear region where the output signal is not directly proportional to the input signal. Therefore, it is still an important step to undergo when nonlinearities are present during the EMC testing.
Having said that, an EMC immunity test measurement consisting on generating a two-tone test signal over a certain EUT was performed at IDNEO – Barcelona (Spain), during the ESR 8 secondment. The measurements were performed in a fully anechoic chamber, using two signal generators Rohde & Schwarz SMY 01 9kHz – 1GHz, a Power Splitter RVZ – 800.6612.52 2.7 GHz to combine the output of the two signal generators, a power amplifiers AR 250A250A 250 watts and a log periodic antenna. The EUT in this experiment is a SONY TV.
The first analyses were performed, considering the test with only one-tone (continuous sine wave) at 227 MHz. The input power levels were increased from -40 dBm up to -16.4 dBm. During this interval, the TV screen became less colourful and the image starts to change to black and white, until the screen completely turns off at -16.3 dBm.
Further, considering the two-tone interference of 227 MHz and 228 MHz, the EUT presented huge amounts of interference in the screen as the amplitude of both tones increases from -30 dBm up to -28 dBm. Applying both tones at the same time, as depicted in Figure 5, the screen presents image problems slightly different from the effect shown in Figure 4. In both cases, the same amplitude levels were applied. When the amplitude level of each tone is -16.4 dBm, the screen completely turns off.
The purpose of this experiment was to verify the output signal that is expected to be generated to create a certain type of interference, in this case, two-tone signals at 227 MHz and 228 MHz. But, instead, according to the Figure 6, there are other frequencies other than the fundamental frequencies that we wanted to test the EUT. As explained before, these are called IM products, which, in this case, arise in the frequencies 2F2 – F1 = 229 MHz and 2F1 – F2 = 226 MHz. In this example, the two-tones interference generates an E-field strength of approximately 4 V/m, which cause the TV SONY screen to completely turn-off.
The analysis of the impact of IM products in an EMC immunity test measuring starts here. If to generate an E-field of only 4 V/m, it is possible to visualize in Figure 6 the generation of IM products if a certain level, that perhaps in this case could not be considered as a hazard. However, if a higher E-field strength is desired to be generated, therefore, the amplitude level of those IM products could considerably impact the final results of the immunity levels. For instance, according to the IEC 61000-4-3 (2020) [1], the test field strength is defined according to the characterization of electromagnetic environment, as shown in Figure 7.
The idea of performing these measurements is a proof of concept on how the nonlinearities generated by power amplifier can affect the final expected measurements. In this experiment, for instance, the EMC immunity level of the EUT is expected to be measured considering only the two frequencies, which could represent a specific EM environment with two different sources of interference. However, as demonstrated, other frequencies are also appearing in the spectrum that, in the end, can compromise the final measurement of the immunity level of the equipment.
As aforementioned, in the realization of an EMC immunity test, it is required to determine the frequencies and power levels that affect the normal operation of an EUT. This step is generally performed applying a rule based approach, more related to the application of harmonised EMC standards. However, considering the complexity of actual EM environment, a risk based approach is more appropriated. The risk based approach is based on the phases of EMC management, control, implementation, validation and verification plan. In the EMC management phase it is defined the EM threats that could be found into the place where the product is installed, for instance, a complex environment with several sources of disturbances [5].
In fact, the EMC immunity test has to accommodate these requirements at the same time it needs to be repeatable and reliable. After all this discussion, here is where the question from the title can be debated, or at least give another possible solution to overcome problems generated by amplifier nonlinearities. And that is the final purpose of the PhD research of the ESR 8: Applying a linearization technique called signal predistortion to address and mitigate nonlinearity issues that can arise during the amplification process in an EMC immunity test. It is useful to reduce distortion, preserving signal integrity, meeting regulatory standards to ensure product compliance and help to control and reduce intermodulation distortion.
References
[1] I.E. Commission, Electromagnetic compatibility (EMC) – part 4 : Testing and Measurement Techniques – section 3 : radiated, radio-frequency, electromagnetic field immunity test, Geneva, Switzerland, 2020.
[2] Duffy, Alistair & Orlandi, Antonio & Armstrong, Keith. (2010). Preliminary study of a reverberation chamber method for multiple-source testing using intermodulation. Science, Measurement & Technology, IET. 4. 21 – 27. 10.1049/iet-smt.2009.0008.
[3] J. Li, S. Ma, Q. Liu, Z. Gong, H. Tian and S. Jin, “Analysis of interference caused by intermodulation in multi-tone radiated immunity tests,” 2018 IEEE International Symposium on Electromagnetic Compatibility and 2018 IEEE Asia-Pacific Symposium on Electromagnetic Compatibility (EMC/APEMC), Suntec City, Singapore, 2018, pp. 660-663, doi: 10.1109/ISEMC.2018.8393863.
[4] L. Devaraj, Q. M. Khan, A. R. Ruddle and A. P. Duffy, “Comparing Simulated Impact of Single Frequency and Multitone EMI for an Integrated Circuit,” 2021 13th International Workshop on the Electromagnetic Compatibility of Integrated Circuits (EMC Compo), Bruges, Belgium, 2022, pp. 107-111, doi: 10.1109/EMCCompo52133.2022.9758612.
[5] W. Ardiatna, D. Mandaris, A. N. Bakti, S. W. Hidayat and F. Leferink, “EMI risk analysis via dedicated evaluation of the susceptibility of medical devices,” 2018 IEEE International Symposium on Electromagnetic Compatibility and 2018 IEEE Asia-Pacific Symposium on Electromagnetic Compatibility (EMC/APEMC), Suntec City, Singapore, 2018, pp. 205-209, doi: 10.1109/ISEMC.2018.8393767.